5M daily-LLM developers by 2027 is the addressable wedge

The claim

The addressable wedge is the developer who runs an LLM prompt every working day. That cohort was ~2M paid seats in 2024 and is on a ~2.5x trajectory through 2027 — a ~5M-seat target market by then.

The evidence

GitHub Copilot disclosed 1.3M paid seats in November 2023 and 1.8M in early 2024 (GitHub investor materials). Cursor reported “over 360k” paying users in mid-2024 (Cursor public memo). Anthropic and OpenAI direct-API developer counts are not separately disclosed but Stack Overflow’s 2024 Developer Survey reported 76% of professional developers using or planning to use AI tools — a ceiling, not a floor.

Daily-cadence usage is the qualifying filter. At an industry average of $20/seat/month, a 1% capture of a 5M-seat wedge is $12M ARR — Series A-shaped. A 5% capture is $60M ARR — growth-stage-shaped.

The comparable

JetBrains reached 12.8M paid users by 2023 from a single-developer wedge (JetBrains state-of-the-industry report). Postman crossed 30M registered users (Postman disclosure, 2023). Both started as single-developer tools, both grew without a team-first GTM. The pattern is: solve the daily individual workflow, the team product follows the user.

What we ask for

A diligence conversation grounded in seat-count math, not LLM-token math. The right benchmark is JetBrains and Postman’s penetration curves over 2008–2018, not the API-spend curves the LLM platforms publish.

pie
  title Daily-LLM developer wedge composition (2027 estimate, 5M seats)
  "Copilot daily users" : 40
  "Cursor and IDE-native daily users" : 25
  "Direct-API daily users (Claude / OpenAI)" : 20
  "Open-model self-hosted daily users" : 15